Another one bites the dust; Clinical trial for prostate cancer prevention by selenium and vitamin E halted

I've written here before about nutritional supplements. Specifically, I've expressed my dismay at the double standard, codified into law in 1994 in the form of the DSHEA. This particular bit of truly awful law in essence took away the power of the FDA and FTC to regulate dietary supplements, except under certain rather narrow conditions. In essence, if a supplement manufacturer is careful to keep the claims for a supplement from being too specific, the FDA is virtually powerless to regulate the supplement because the law defines dietary supplements as "food." So, in other words, vague, meaningless claims, such as that a supplement will "boost the immune system" are hunky dory, and supplement manufacturers only get into trouble when they say things like, "This supplement can cure cancer." Worse, the FDA is overburdened and apparently doesn't consider policing these supplements, which are a multibillion dollar a year business even when their manufacturers do cross the line. Indeed, only recently, after a long period of doing little, the FDA actually cracked down hard on some notorious quacks. I roundly applauded the action, but I couldn't help but ask what too the FDA so long.

One of the reasons I rail against the current law is that it places supplements in an exalted place, almost immune from regulation, despite the fact that snake-oil peddlers make the most outrageous claims for what they can do for your health. But there's an even more insidious effect of the current law. Supplements could potentially have health effects, both beneficial or detrimental. They are substances that our bodies need, but in the optimal amounts. At a conference in New York today, I was listening to a discussion about vitamin D and breast cancer (something I'll have to write more about soon), and that point was made again and again: Just because some of something is good does not mean more of it is better. In any case, we can't know what health effects various vitamins and supplements might have and at what doses if we don't do the research. Unfortunately, thanks to the lax regulation of these supplements and the dubious practitioners making all sorts of dubious claims for what they can do, vitamins and supplements as potential health modulators have undeservedly taken on the taint of quackery. They've been, in essence, "ghettoized." Instead of being studied with funding from Institutes at the NIH that do rigorous peer review, studies of the health effects of supplements have fallen under the purview of the woo-friendly National Center for Complementary and Alternative Medicine (NCCAM).

Fortunately, some good science is being done studying supplements. Unfortunately, one such trial, the Selenium and Vitamin E Cancer Prevention Trial (SELECT), has had to be halted because the supplement showed no benefit and showed a hint of risk:

WASHINGTON - The government is stopping part of a major study of whether vitamin E and selenium prevent prostate cancer -- because the supplements aren't working and there's a hint of risk.

More than 35,000 men age 50 and older have been taking one or both supplements or dummy pills as part of a study called the SELECT trial.

But the National Cancer Institute announced Monday that they will be getting letters in the next few days telling them to quit the pills. An early review of the data shows neither supplement, taken alone or together, is preventing prostate cancer.

Of more concern, slightly more users of vitamin E alone were getting prostate cancer -- and slightly more selenium-only users were getting diabetes, the NCI said.

The increase in prostate cancer among vitamin E users and the increase in diabetes among the selenium-only users was not statistically significant.

According to the National Cancer Institute website:

"SELECT was always designed as a study that would answer more than a single question about prostate cancer," said Eric Klein, M.D., a study co-chair for SELECT, and a physician at the Cleveland Clinic. "As we continue to monitor the health of these 35,000 men, this information may help us understand why two nutrients that showed strong initial evidence to be able to prevent prostate cancer did not do so."

SELECT was undertaken to substantiate earlier, separate findings from studies in which prostate cancer was not the primary outcome: a 1998 study of 29,133 male smokers in Finland who took vitamin E to prevent lung cancer surprisingly showed 32 percent fewer prostate cancers in men who took the supplement, and a 1996 study of 1,312 men and women with skin cancer who took selenium for prevention of the disease showed that men who took the supplement had 52 percent fewer prostate cancers than men who did not take the supplement.

Based on these and other earlier findings, in 2001, men were recruited to participate in SELECT. They were randomly assigned to take one of four sets of supplements or placebos, with more than 8,000 men in each group. One group took both selenium and vitamin E; one took selenium and a vitamin E placebo; one took vitamin E and a selenium placebo; and the final group received placebos of both supplements.

This is normally how clinical science works. Based on scientific plausibility plus preclinical scientific studies in vitro and in animal models, smaller, preliminary clinical trials are performed and generate evidence suggestive that a given therapeutic or preventative intervention is efficacious. However, the weaknesses and drawbacks in such study make it impossible to conclude with sufficient confidence to change clinical practice, and the question is considered important enough that it needs to be answered. Hence, a randomized, double-blind clinical trial is carried out. Often, more than one are carried out asking essentially the same question. Data from these trials then drive practice. All too often in complementary and alternative medicine" (CAM), preliminary trials done over and over and over are the "evidence" that is used to justify woo. Such was not the case for the SELECT trial.

Indeed, the SELECT trial was designed to go the right way: testing in a large randomized trial whether the results of preliminary trials held up. SELECT accrued 35,000 subjects and is costing $114 million to carry out. If a CAM booster ever tells you that "there's no money" to study supplements, point him to the SELECT trial, which was incredibly expensive--serious bucks. Indeed, I'm not entirely sure whether the question of whether vitamin E and/or selenium can prevent prostate cancer is important enough to warrant spending that much money on it, although if it hadn't been done then we might still be recommending that men take these supplements and providing for them no benefit.

But why did the government stop the trial? Ethics demanded it. One rule of clinical trial design and monitoring is that trials in which the therapeutic agent may be causing harm must be stopped early. In this case, there was no evidence of benefit from the vitamin E or selenium supplements in terms of reducing the risk of prostate cancer. Couple that with the trend, albeit a not statistically significant trend, towards harm in the groups of men taking the supplements. Ethically, it was a no-brainer to stop the trial. Unfortunately, the result was yet another negative trial regarding supplements and health benefits claimed for them. It's a shame; as a man who can no longer avoid the label "middle aged," I would have been quite happy if the results did show that vitamin E and/or selenium prevent prostate cancer. Who wouldn't be happy if a cancer could be prevented just with a pill?

Sadly, it was not to be, but at least we know with a high degree of certainty that vitamin E and selenium appear not to prevent prostate cancer.

More like this

@ Joseph C.,

I must be psychic! Shortly after starting to read this I said to myself "Self, the quacks are going to rationalize-away the results." Thanks for confirming my hunch.

For folks unaware of this propensity of woomeisters, get James Randi's book "Flim Flam."

What probably took the FDA so long to police this would likely be the the passion for "deregulation" we've been suffering the last 20 or so years. Where the "government is bad" folks cant get regs repealed, they cut funding for enforcement. This has been done to workplace safety, environmental enforcement, tax collection, and other areas

Another possibly naive question:

You mention above "Based on scientific plausibility..." and you've talked about that before. While I agree completely that it's silly to spend money to study things that have no possible mechanism of action other than supposed supernatural ones, I'm not sure I really understand why scientific plausibility and/or a possible mechanism are requirements for starting a trial. Would you then argue that, if the researchers had no idea HOW selenium or vitamin E might reduce prostate cancer risk, they should not have been allowed to pursue the trial? (Implicit in your statement is the idea that they did have a plausible mechanism already, even if it wasn't mentioned in the press releases you quote.)

That seems backwards to me, and contradicted by a lot of examples I've heard about. For example, it wasn't known for a long time how aspirin reduces heart attack risk (as far as I understand, anyway), but there were still lots of studies done to quantify that reduction in risk and show that aspirin does, in fact, reduce heart attack risk. Similarly, as far as I know there isn't really any understanding or mechanism known for why excess body fat might cause diabetes, but many studies are conducted to try to prove that excess body fat does cause diabetes.

As I said, I completely agree that things with no possible realistic mechanism (such as homeopathy or crystal healing or whatever) shouldn't be studied...but why is having a plausible mechanism a requirement for clinical trials to proceed? Or am I misunderstanding what you mean by scientific plausibility?

@ Dr. Kate,

I look forward to Orac's more articulate response. My take is- it is not just the impossibility of a method (e.g., homeopathy); but what is the evidence that suggests a large trial is worthwhile.

Since, in this case, this huge trial showed no benefit, one wonders what properly-done pilot studies justified this massive study. It seems more likely that it was merely a preponderance of anecdotes. Anecdotes may make a pilot study plausible; but not something as massive as this.

It's a shame; as a man who can no longer avoid the label "middle aged," I would have been quite happy if the results did show that vitamin E and/or selenium prevent prostate cancer.

Definition "middle-aged:" That point in a man's life when "small and soft" becomes good news.

And let me observe, with a few more miles on the odometer than you have, that it's a hoot.

By D. C. Sessions (not verified) on 30 Oct 2008 #permalink

I'm not sure I really understand why scientific plausibility and/or a possible mechanism are requirements for starting a trial. Would you then argue that, if the researchers had no idea HOW selenium or vitamin E might reduce prostate cancer risk, they should not have been allowed to pursue the trial?

Life is short and resources are limited. Although we might not know the exact mechanism of action, there's no a priori reason to doubt that one exists, so investigation into well-established empirical observations is well warranted.

Likewise investigations into strongly indicated theoretical effects, at least to get a preliminary empirical confirmation.

The ugly comes up when the only suggestion is for something with no theoretical support and a dearth of solid empirical support (e.g. chelation for atherosclerosis.) At best full trials for something like that would be a waste of money, and at worst horribly unethical (e.g. the Gonzales regimen for pancreatic cancer.)

By D. C. Sessions (not verified) on 30 Oct 2008 #permalink

Dr. Kate,

Your line of thought is good, but your examples don't actually counter "prior plausibility." Aspirin's basic mechanism of action was first proposed in the 1960s, about the same time that clinical trials for aspirin as a means of reducing heart attacks and strokes came about. I'm not really sure from a historical standpoint what we knew about stroke risk factors in 1960, but warfarin was first used for anticoagulation in the 1950's--which suggests to me that we had some idea that anticoagulants/antiplatelet drugs could help reduce stroke risk. The precise mechanism of action of aspirin wasn't discovered until 1971, but at that point I imagine it was mostly a confirmation of what we already knew.

Likewise, body fat isn't simply inactive mass, it's metabolically active and has secretory functions. It's possible that body fat increases the risk of diabetes because of the dysregulation of adipose tissue-produced hormones, increased adipose tissue reducing sensitivity to insulin and incretins, etc. But the idea that there's no prior knowledge to make a connection there isn't entirely true.

Before a clinical trial proceeds there really ought to be some hypothesis as to why a particular treatment might work, otherwise we're just shooting blindly. It's true that some treatments have been discovered serendipitiously, but you can't rely on that method, and there's only so much time and money available for research.

~N.B.

Quoth NaturalNews:

Vitamin E and selenium work synergistically to support immunity, antioxidant defense systems, and cancer prevention. No bogus study will ever discredit these simple facts.

It's nice to see inadvertent honesty: "Don't confuse me with facts, my mind is made up."

By D. C. Sessions (not verified) on 30 Oct 2008 #permalink

i'm confused. what is naturalnews talking about when they contrast "chemically-derived vitamin E (dl-alpha tocopherol acetate)" with naturally-derived vitamin E?

If you're interested, Kate, there is a stack of posts on prior probability at sciencebasedmedicine.com. Among the reasons that plausibility is an issue is Bayesian stats. The likelihood of a positive result being false is very high the more implausible the starting hypothesis.

NB wrote "Before a clinical trial proceeds there really ought to be some hypothesis as to why a particular treatment might work ..."

No, a strongly supported hypothesis would be valuable; but, good, clinical evidence is also valid (even, overwhelming). Aspirin exemplifies that point- it was in use for decades before any idea of its pharmacological effects emerged.

People take heart as progess on the learning curve has been accomplished. Take Mercury. With that, woo masters and their progeny ignored the the dose makes the poision principle. Yet with Selenium, a far more toxic substance, they recognize that it can be safe in small doses. Even useful.

@ D. C. Sessions,

It's really entertaining to see watch the intellectual gymnastics of Mike Adams and pals unfold. They walk the impossible tightrope of trying to use science to lend credibility to their moon bat beliefs while simultaneously tearing down any science that threatens their fragile weltanshauung. They bash science for being reductionist while proclaiming "simple facts" out of staggeringly complex and poorly understood biochemical interactions. It's hilarious.

Especially hilarious is how these mental midgets don't comprehend "the simple fact" that a molecule is a friggen molecule, synthetically or naturally synthesized.

By Joseph C. (not verified) on 30 Oct 2008 #permalink

I love how in the linked naturalnews article, it was not only the wrong Vitamin E, it was the wrong synthetic Vitamin E, and it was even the wrong EVIL GERMAN synthetic Vitamin E. I'm surprised he missed the obligatory Cylon-B reference.

Joe,

Your point is valid, but if you've got clinical data supporting something's efficacy you should be able to come up with some hypothesis explaining how it might work.

@ N.B. wrote "if you've got clinical data supporting something's efficacy you should be able to come up with some hypothesis explaining how it might work ..."

That is a nice philosophy; but it is not born out in practice. Aspirin is a perfect example of a drug that was known to work decades before we could understand why.

i'm confused. what is naturalnews talking about when they contrast "chemically-derived vitamin E (dl-alpha tocopherol acetate)" with naturally-derived vitamin E?

They're talking the same nonsense all chemistry-ignorant people do. They think there's a difference between a molecule made in a synthetic process and the same one extracted from "natural" stuff, the "natural" one, of course, being somehow magically "more energetic" or "immunity-boosting" or whatever. They like to get out the molecule's real (meaningful to chemist, that is) name, which is so much more chemical(read, evil and poison)-sounding. Vitamins generally have several forms, which might confuse matters. One of the forms of Vitamin E (yes, even the magical, natural form) is alpha-tocopherol. Every time someone talks about "chemicals" that way, I just snort in my coffee now.

Dr. Kate:
Similarly, as far as I know there isn't really any understanding or mechanism known for why excess body fat might cause diabetes, but many studies are conducted to try to prove that excess body fat does cause diabetes.

To go into this subject with a little detail...
It has been empirically established for centuries that excess body fat was related to diabetes. It was about 25-30 years ago that people noticed that there was a difference in susceptibility to diabetes [type 2] that was related to the distribution of body fat - "apples" were highly correlated, "pears" were not. This progressed to actual measurements of subcutaneous v. intra-abdominal fat, so that the empirical observations about fat and diabetes were well established by about 20 years ago. The next steps were the recognition of the differences in anabolic/catabolic differences between the two fat types followed by the discovery of differences in endocrine and other activities over the last 20 or so years. There has been a steady progression of knowledge about the nature of IA fat, pancreatic islets, target cell interactions, blood lipids, etc. so that the "whys" are fairly well understood by now, with more details needed, of course. However, there is a lag time between what research has discovered and what becomes general health/medical knowledge and how much of that knowledge is pertinent to the medical practitioner. There's always a lot more out there than you really want to know.

By natural cynic (not verified) on 30 Oct 2008 #permalink

Tangentially on topic, has anybody heard of a supplement called "Body Balance" by the company Life Fore International? It's apparently only available by direct sale and some friends who are reps around here have been trying to sell me on it. They claim it's "phytonutrients" in juice form which is supposedly better for you than the pill vitamins you get in the store.

Thing is, the bottles are freakin' expensive. Just curious if anyone's woo-meter goes up at this, if I should save my money?

By Brian Hodges (not verified) on 30 Oct 2008 #permalink

An amendment: apparently the pytonutrients are derived from "sea vegetables" for whatever that's worth.

By Brian Hodges (not verified) on 30 Oct 2008 #permalink

@ Brian,

http://php.lifeforce.net/body-balance.php

Looks like very standard woo. Not the full-on Mike Adams with-this-magic-pill-cancer-can't-grow-in-the-body woo, but rather a more gentle strain.

My favorite part was:

Furthermore, the chemical balance of the human body is similar to that of sea water.

They, of course, fail to define what the hell they mean by "chemical balance". And I don't recall ever being tested on any such concept in the several chemistry classes I took in college.

Then we have the standard vague health claims:

Ever since I started using Body Balance, my health has continued to improve.

Of course, this gentleman fails to say how his health has improved. Nothing but vagueness here. Not to mention this sort of suggests that his health was already improving before he started taking this crap.

I'll never stop taking Body Balance - it's the best!

Wow, I'm sold!!!

Save your money, bro.

By Joseph C. (not verified) on 30 Oct 2008 #permalink

Sea vegetable confused me at first, until a bit of googling showed it's the woo name for algae. Extract of sea vegetable sounds so much better then mashed up sea weed.

Joseph C, the "balance" is their way of saying "look! cytosol and seawater have the same concentration of sodium!" 150mmol, IIRC.

I like seaweed salad. Does that count for anything?

By Rogue Epidemiologist (not verified) on 30 Oct 2008 #permalink

And the woo just keep on coming from our friends at Natural News.

When one examines the study itself, it becomes apparent what Dr. Messing was referring to.
Instead of using the natural form of Vitamin E considered most beneficial for health, d-alpha-tocopherol acetate, they used dl-alpha-tocopherol acetate. The "d" designation in front of the "alpha" indicates that the products are derived from natural sources such as vegetable oils or wheat germ. A prefix of "dl", such as in dl-alpha- tocopherol, shows that the vitamin has been synthesized from a petroleum base.

The ignorance of orgo here is astounding. It seems these Nobel candidates don't even know what racemization is. They probably don't even know that the d-isomer is still present in the racemic mixture. Nor do they understand that using a "petroleum base" as a starting material is irrelevant since there's this whole thing called "purification".

Jesus Christ these people are stupid.

By Joseph C. (not verified) on 30 Oct 2008 #permalink

The ignorance of orgo here is astounding. It seems these Nobel candidates don't even know what racemization is. They probably don't even know that the d-isomer is still present in the racemic mixture. Nor do they understand that using a "petroleum base" as a starting material is irrelevant since there's this whole thing called "purification".

Jesus Christ these people are stupid.

Not at all. The fundamental test is very simple: does their patter work to separate the marks from their money?

It does. Thus they aren't stupid.

By D. C. Sessions (not verified) on 30 Oct 2008 #permalink

Not at all. The fundamental test is very simple: does their patter work to separate the marks from their money?

It does. Thus they aren't stupid.

Very, very true, but some of the best woo peddlers actually believe their own stupid. Adams seems to fall into this category. Didn't you read the part about him being an "independent journalist"? :)

By Joseph C. (not verified) on 30 Oct 2008 #permalink

"In this case, there was no evidence of benefit from the vitamin E or selenium supplements in terms of reducing the risk of prostate cancer. Couple that with the trend, albeit a not statistically significant trend, towards harm in the groups of men taking the supplements. Ethically, it was a no-brainer to stop the trial."

Why is a statistically insignificant result a no brainer? Also, shouldn't they look for non-targeted health benefits before canning the trial?

Is there any indication that they didn't look for "non-targeted health benefits"? After all, there was a finding of a trend towards more diabetes cases in the selenium users; presumably fewer diabetes cases would have been noted as well.

A question: did the SELECT trial last long enough to really tell us "with a high degree of certainty" that vitamin E and selenium don't protect against prostate cancer?

Incidentally, Mike Adams missed a bet in his diatribe against synthetically-produced vitamin E. He did note that it was derived from coal tar, but failed to point that mercury (MERCURY!) is a constituent of coal tar.

http://www.sciencedirect.com/science?_ob=ArticleURL&_udi=B6V3B-47XW574-…

Obviously the government is now concealing the facts about an epidemic of autism in the men who took part in this study.

By Dangerous Bacon (not verified) on 30 Oct 2008 #permalink

Why is a statistically insignificant result a no brainer?

A "Statistically insignificant" result over 35,000 people is still a lot of people and when there is absolutely no benefit being given then the Risks outweigh the Rewards.

Also, shouldn't they look for non-targeted health benefits before canning the trial?

No. the trial was carefully designed to look for certain results, changing the focus of a trial halfway through is a no no. Whilst they may find some interesting correlations along the way the data produced would probably be unreliable, but a good jumping off point for new research.

Thanks for all the answers. (I had forgotten about Bayes--and you're right, that does make sense.)

Thing is, the bottles are freakin' expensive. Just curious if anyone's woo-meter goes up at this, if I should save my money?

My take on nutrients is the best, cheapest source for them is called 'food'. There is no need for you to take any kind of supplement if you have a healthy, normal diet and no diagnosed illness. The only nutriments which might pose a problem for us, ie vitamin D and iodine, are already added in food (milk and salt). Vitamin supplements are at most useless and at worst harmful (especially if you take things like iron or folate without a diagnosed deficiency).